Pylimitics

Chinese large language models like DeepSeek are smaller and more efficient than the centralized commercial behemoths from OpenAI and Anthropic, et al. They’re also open source, so you can use and implement them however you want. They’ve been released to the world and they’re free and open.

The US response is to create a new bunch of sales reps. Never mind actual performance and fitness for needs, the point is talking people into something whether they want it or not. By the way these new sales reps are expected to be volunteers, so they don’t even get commissions.

Isn’t it interesting that at least in AI tech, the US is now the champion of centralized, monolithic systems while China is exactly the opposite. And this is a big change in how the US operates. As Paul Krugman points out, “The people guiding U.S. trade policy between 1934 and the 1960s or so believed that increasing global commerce would promote peace and prosperity, but they didn’t expect to get open world markets by lecturing people about the virtues of free trade.” (emphasis added). The creeps in charge now seem to think lecturing and hectoring is the way to go. Possibly because they don’t have the wit to formulate anything rationally.